Chris Blattman et. al. report long-term RCT results for a program of cognitive-behavior therapy (CBT) + cash transfers for criminally-involved men in Liberia. Quick take: High-quality RCT finds sizable reduction in anti-social behavior (e.g., drug-selling, robbery) at 10-year follow-up.

Program:

• This was an 8-week group CBT program - Sustainable Transformation of Youth in Liberia - focused on developing skills for planning, goal-setting & decision-making; controlling emotions/impulses; & fostering a noncriminal self-image. It also included a one-time $200 cash grant.

Study Design:

• The study randomly assigned 999 criminally-involved young men to (i) CBT alone; (ii) CBT + cash; (iii) cash alone; or (iv) control.

Findings:

• The study found, at the 10-yr mark, that CBT + cash significantly reduced anti-social behavior vs the control group (effect size -0.25, p<0.05).

• The effect appears driven mainly by decreases of nearly 50% in drug-selling & thefts/robberies. The effect of CBT alone (vs control) was almost as large and close to statistical significance: effect size -0.20, p<0.10. Earlier study follow-ups had also found sizable effects.

Comment:

• Based on careful review, this was a high quality RCT (e.g., good baseline balance, low attrition). Outcomes were self-reported but many years after the program, lessening concerns about social desirability bias (e.g., treatment group reporting behavior they'd just been taught as desirable).

• Overall, I think very promising. A replication RCT in another location would be valuable to hopefully confirm the result & establish its generalizability to other settings.